On 11/26/23 23:56, Dave Everett wrote:
If it were me, and it's not.
I would carefully replace each wire with a low to the board wire and
double-check everything. I have rushed in to this kind of thing before
and regretted it.
Surprisingly that monstrosity didn't work.
This slightly better but still probably nowhere near what you meant, did
though!
https://github.com/bkw777/NODE_DATAPAC#minindp-512
My first idea for how RAMDSK expects to do 512k support didn't work, but
my next guess did. I have a working 2-bank 512k breadboard!
So now I'm just waiting for the pcbs just to be sure before posting the
gerbers and updating pcbway and making this the main version.
Dave
On Mon, 27 Nov 2023 at 15:10, Brian White <[email protected]
<mailto:[email protected]>> wrote:
I am absolutely certain that every one of these connections is correct.
https://photos.app.goo.gl/mzJbKM3cWJ8AgGic8
<https://photos.app.goo.gl/mzJbKM3cWJ8AgGic8>
This is essentially a MiniNDP (which is essentially a Node Datapac
but with a single large sram), minus the battery stuff, in the form
of all through-hole dip parts on a breadboard.
I think I deduced how the 2-bank 512k rampac is supposed to work by
a combination of studying the single bank 256k datapac and capturing
bus traffic from ram100.co <http://ram100.co> when it tries to
access the 2nd bank.
I already drew up an updated schematic and pcb and ordered some
pcbs, but still I figured since I actually have dip versions of all
the chips, and it's only 5 chips plus the sram, I'd try to
breadboard the whole thing and use that to see if it works first.
Start with a simple copy of the known-working 256k circuit, then
once that is verified, add my guess at the few changes to support
bank1 and 512k.
For some strange reason I hesitate to actually try it now. Might
have something to do with the 3 or 4 errors I already found and
corrected along the way.
--
bkw
--
bkw
